Hi, my name is

Arinjay Singh.

I'm an aspiring software engineer studying Computer Science and Economics with a particular interest in the ethical implications of sociotechnical systems.

Experience

Data Analyst Internship

Neocortex, Fremont, CA

July 2023 - August 2023

Full-Stack Software Developer

Northeastern Electric Racing, Boston, MA

September 2023 - Present

Food Rescue & Recovery Volunteer

White Pony Express, Pleasant Hill, CA

August 2020 - August 2022

Research

As a research assistant under Professor Xiaolin Shi, I explored the various health implications of grandparents co-residing with their grandchildren utilizing self-rated health. More specifically, I utilized the field of econometrics, applying Difference-in-Difference Estimation and Regression Discontinuity Design on 4,000+ grandparent-grandchild pairs pulled from the Panel Study of Income Dynamics (PSID) conducted by the Survey Research Center at the University of Michigan.

Natural Language Processing

Research Assistant to Professor Mai ElSherief

As a research assistant under Professor Mai ElSherief, I am working to utilize the field of Natural Language Processing (NLP) to identify migrant misinformation patterns on social media in the U.S. and EU. This research is funded by the Google Trust and Safety Research Award and aims to create positive societal impact with technology.

Additionally, as part of a secondary project I am developing a framework based on the principles of Data Feminism for evaluating the ethical use of data in current NLP systems. This framework will be used to evaluate the ethical implications of the data used in the research and the potential biases that may arise from it.

Projects

Cyberbullying Sentiment Analysis

Python

Scikit-Learn

Pandas

NumPy

NLTK

TensorFlow

Keras

Git

In this project, I developed a Bi-directional Long Short-Term Memory (BiLSTM) model utilizing Word2Vec dense word embeddings to classify social media posts as cyberbullying or not cyberbullying. The model was trained on a joint sentiment and cyberbullying dataset of 250,000 tweets, achieving a weighted F1-score of 0.81 and an accuracy of 81%. It's relative performance was further confirmed by comparing it to a Logistic Regression model and VADER, a leading sentiment analysis tool.

Personal Website

HTML

CSS

JavaScript

React

Next.js

Tailwind CSS

Git

This is my personal website that I designed with Next.js, styled with Tailwind CSS, and deployed using GitHub Pages. It will serve as a place for me to document my experiences, projects, and skills as I continue to grow as a software engineer.

Information Retrieval Functions

Python

NLTK

Elasticsearch

Kibana

Git

In this project, I implemented various document ranking functions for information retrieval systems such as Okapi TF, TF IDF, Okapi BM 25, and Laplace Language Model. The functions utilized an inverted index of a corpus to rank documents based on their relevance to a given query.

Northeastern Healthy Dining

Python

Scikit-Learn

Matplotlib

Pandas

NumPy

In this project, I trained a Random Forest Regressor model to predict the healthiness of Northeastern dining hall meals based on their nutritional content, utilizing Nutri-Score (WHO-recommended nutritional rating system) to determine the healthiness of the ingredients.

Secure Password Generator

Python

Git

In this project, I developed a secure password generator based on the XKCD method. The XKCD method is a password generation method that originates from a webcomic claiming that people use passwords that are hard for humans to remember, but easy for computers to guess.

Sodoku Solver

Python

Git

In this project, I developed a backtracking algorithm to solve Sodoku puzzles. One implementation simply solves the puzzle, while the other implementation allows the user to play the game themselves and view the solution if needed.

Pong

Python

Git

This is my personal interpretation of Pong, a classic arcade game. It features a single-device two-player mode where players can compete against each other to see who can score the most points.